2021 Nashville AISTech

Applying Several AI Approaches to Improve the LF Steel Quality by Sulfide Capacity Estimation and Slag Chemical Composition Prediction (Room 209 A)

01 Jul 21
9:30 AM - 10:00 AM

Tracks: Digitalization Applications: Software Innovations

Speaker(s): Alex Alvarez, ECON Tech; Maria Luisa Argaez, Data Scientist, ECON Tech; Nelson Enrique Sanchez, ECON Tech

Desulfurization is an important process for steel refining. The standards specify increasingly lower sulfur content. The sulfide capacity index (Cs) is used to describe the slag potential to remove sulfur. A machine learning + LF digital model was developed to predict the slag chemical composition and then used as input into the artificial neural network (ANN) Cs estimation model. The result of ANN showed an R2 equal to 83.26%, MSE of 0.0045, and MAE of 0.0534 using the validation data set. Statistics parameters display acceptable results, showing a robust model capable of generalizing the results for new data entry.